Su Qin's words suffocated the atmosphere in the venue.

It was a silent sigh of indignation—everyone's rationality was snatched away in an instant by the possibility described by Su Qin, and he felt that what he said made sense.

But subconsciously, everyone can't help but reject and resist, because the topic of "the combination of the Matthew effect and the Internet" is too heavy - everyone who can't achieve the first place in a certain subdivision seems to become worthless scum. But how many can there be in the first place? No matter how detailed the field of judgment is, I am afraid that 90% of the people in this world will become scum in the future, right?

While everyone agreed with Su Qin, they couldn't help but look forward to Feng Jianxiong's rebuttal more than before.

Feng Jianxiong spoke appropriately: "The head content providers in each specific segment will be magnified by the Matthew effect, and the second place, even if it is only 5% worse than the first place, will become worthless - this judgment, on the whole, is no problem." ”

He first summarized the other party's description, highlighted the key points to be analyzed, and then changed the conversation and focused on dissection.

However, please note that this assumption of yours only exists in an ideal state. In fact, in real life, there are only a very small number of people who can rely on a single evaluation index to achieve the first place in a certain sub-field - we can assume that they are all 'professionals', and the proportion of the value and contribution they create to society is also insignificant compared to the total value created by human society.

Even if the technology of dissemination and reproduction is more advanced in the future, so that the physical cost of dissemination and reproduction will be reduced to zero. At that time, human beings naturally had sufficient individualized and differentiated needs to be discovered, and they became more and more picky - at this time, those who created mainstream value for society were still individualized and integrated generalists......"

Feng Jianxiong said this, and finally summarized the summary of this passage.

Su Qin, who had been waiting patiently, finally seized the opportunity and decisively launched a counterattack: "Your theory is just empty words. On the one hand, you admit the coexistence of the two types of value creation systems, and on the other hand, you can't cite any global data evidence, and you can only assert that one of them is the mainstream, which is too irresponsible, right?"

Feng Jianxiong did not give in: "There are indeed no global statistics at present - it's not that I don't want to collect them, but based on the existing technology, it is impossible for any country's statistical department to complete this kind of statistics." But as debaters looking to the future, we should learn to make the right outlook by looking at concrete examples and trends.

In the modern technological society, is it still common to be able to rely on time and energy to slowly sharpen the industry of excellence? Perhaps, you can learn from Jiro Ono to make sushi for 60 years, or learn Tetsuya Saotome to fry tempura for 50 years, and finally make cooking and craftsmanship, and improve a skill from 98 points to 100 points - but how many such industries are there?

AT PRESENT, WE ARE SEEING MORE OF THE PRACTITIONER SKILLS REQUIRED BY AN INDUSTRY, ITERATING AT AN UNPRECEDENTED AND RAPID FREQUENCY, AND A PERSON WHO HAS LEARNED C++ MAY HAVE TO SWITCH TO JAVA WITHIN THREE YEARS. A StarCraft player who has worked hard to practice micro-exercises, because of a revision by Blizzard, it is not even worth a fart after a few years of hard training. In such an era, you let most people waste their time making tea and cooking for 50 years, enlightenment and refinement, isn't it a mistake. ”

Su Qin immediately ridiculed him: "The opponent's defense friend is secretly changing the concept again - just now we have already said in our argument that any pre-skill required to learn new skills is a quality that a professional should have."

C++ is meaningful for learning JAVA, and a person who has learned C++ is definitely faster than a pure newcomer who doesn't know any programming language, and can better grasp the essence of the idea. A person who has learned the complex transformation of high number line generations must be better at learning the principle of automatic control than someone who has not learned high mathematics. And in the mouth of the opponent's defense friend, this kind of accumulation and precipitation has become worthless, isn't it ridiculous?"

The confrontation between Su Qin and Feng Jianxiong made the scene stalemate for a while.

Several other debaters inevitably interjected occasionally during the process, so that the main force could have some rest.

Often, Yu Meiqin and Nan Xiaoyuan just said that "the speed of modern scientific and technological progress has made most cutting-edge creative workers have to learn for life, and it is impossible to expect to hone their skills for a lifetime";

On the other side, Mi Suli and Niu Zeqian used lines such as "lifelong learning does not mean giving up specialized stock knowledge and starting from scratch completely".

As the confrontation deepened, the core offensive and defensive points of the topic were gradually focused, and there was no room for detours.

After Feng Jianxiong relieved himself, he launched a general attack on the final fortress.

Of course, he did not hesitate to use his foreknowledge of the future to cheat - there is nothing to be ashamed of. After all, he knows better than Su Qin what kind of people the future society needs, and what kind of corner the progress of science and technology will force the evaluation system of human values.

"May I ask Senior Su, have you heard that Google has invested in a research institute of a professor at the University of Toronto named Jeff Hinton? If you haven't heard of it, I can tell you now that in the past two years, he has invented a kind of artificial intelligence route called 'deep learning algorithm'.

Although there are not many applications at present, there are already companies in China based on this algorithm idea, making a new intelligent pinyin input method that can count the frequency of each person's typing words. And Google has also used this algorithm idea to start teaching robots to play Go, compose routine-style tunes, and even compose poems......"

Su Qin felt a hint of danger, but immediately questioned the relevance of the debate topic, trying to stop Feng Jianxiong from casually pulling away from the topic: "So what? This has nothing to do with today's debate. ”

"It's related, because I can show evidence that based on this technology, if humans in the future want to compete with machines by 'slow grinding', then humans have no chance of winning. Google's chess-playing robot will take at least five years to be built, but they have already laid the logical foundation for 'Monte Carlo tree training'.

In the future, robots that have been set with a single goal will be able to evolve themselves by playing 1 million games of Go against themselves every night and counting the changes in the winning rate of each different move and response. At that time, the human Go world champion will probably be as miserable as when Kasparov fell to IBM's 'Deep Blue' in '97. ”

"It's hard for you to imagine this state at the moment, because at the moment we and the computer seem to be fighting against Warcraft and DOTA's man-machine battle - even if the most advanced difficulty is set, the computer only relies on more HP, attack power, and money collection speed to gain an advantage over humans.

In addition, there may be the computer's always wonderful micro-operations - when the computer is playing Warcraft, when it sees that a soldier's HP drops too quickly, it will automatically drag it back and briefly leave the battle line until the opponent's melee unit turns to attack other ones with more health. This kind of operation requires a long time of training for human masters to do, and it is difficult to operate multiple lines at the same time due to the limited hand speed. But for the computer, even if there are 100 places on the map fighting at the same time, it can also manipulate 100 battlefields at the same time.

But what about the future? Computers may be AI that crushes human players directly at the tactical or strategic level - as long as the deep learning AI used to play games is set with the sole goal of 'winning by all means'. At that time, the machine can first learn a training set, obtain basic strength, and then use the Monte Carlo tree-like training structure to fight tens of millions of rounds left and right, keeping in mind the changes in the trend of winning and losing - at this time, what is the value of those human players who hoard skills?

In other words, even if they are still valuable, and there are human viewers who want to watch them play, it is not because 'they play better or worse than the machine' – but because they are still human beings, and because they are different from the machine's 'human' attributes, they have won sympathy points. They are relying on their own abilities beyond their own expertise to gain their own value and dignity. ”

Has the Monte Carlo tree-like training model been proposed on this earth today? Perhaps, we should be able to see the difficult and obscure academic monographs.

So Feng Jianxiong is not a rumor.

As for whether Google has started to do this and use this thing for commercial research, who knows.

This thing is a trade secret, so as long as it exists academically, Feng Jianxiong has nothing wrong with saying this.

In the future, when Google really does it, human beings will only marvel at Feng Jianxiong's foresight and subtlety.

However, Su Qin will definitely not give up.

He also can't be positive and hard-hearted to discuss black technology with Feng Jianxiong - as long as he shows his moderate disdain.

"Hehe" Su Qin snorted coldly, pausing for a few seconds to attract the audience's attention, "When can even the random speculation in science fiction be used as evidence for debate?

Since ancient times, the performance of machines has replaced human skills, and since ancient times, it has been nothing more than those jobs that are simply repetitive or the pursuit of precision. Perhaps the value of those human 'professionals' who pursue 'precision, strength, and agility' will indeed be eliminated and replaced by machines. But what about original, research-based work? Machines can only copy human behavior patterns, and can they still actively create them?"

Feng Jianxiong smiled and took out a copy of Wired magazine.

It was just published in recent months, and it should be all cutting-edge scientific and technological achievements.

In 06, in the embryonic period of less than two years after the birth of deep learning algorithms, it was not easy to find some ironclad evidence at the level of "alpha dog".

Not entirely absent, though.

AT LEAST FACEBOOK HAS COME UP WITH A PROTOTYPE OF SOFTWARE THAT CAN REPLACE HUMAN ARTISTS IN RETOUCHING. After a few years, this kind of thing will proliferate, and then it will become a smart version of the "Meitu App" that can automatically edit photos.

Moreover, it is not the first time that Feng Jianxiong has mentioned "deep learning artificial intelligence" in the school's debate competition - he has already mentioned it once in last year's competition, and later was interviewed by Ding Lihui, the female anchor of the school station, to do a special prospect program to discuss with students what skills and attributes are needed for future talents.

It's just that every time Feng Jianxiong mentions artificial intelligence, he can say a lot of new tricks from different entry points and angles.

Therefore, when I mentioned it again today, whether it was Su Qin on the opposite side or the judges on the field, it was somewhat easy to judge whether what Feng Jianxiong said was true.

Holding the magazine in his hand, Feng Jianxiong asked: "May I ask the other party, is the work of a painter creative labor - note, I am asking about a painter, who paints for artistic purposes, not purely for the purpose of 'painting like', so don't refute me with photographic techniques." ”

"The work of the painter is, of course, original. Su Qin didn't think there was anything wrong here.

Feng Jianxiong continued to press: "Then, isn't the work of using PS to help humans retouch and artistically process photos?"

Su Qin felt a hint of danger, and after hesitating for a moment, he gritted his teeth and affirmed: "Of course it is, but does it matter?"

"I'm sorry, I think I have to tell you - as long as a job is judged by a single criterion, the goal is objective and determined, then in the future machines can replace and eliminate humans. ”

He flipped the magazine to a folded page and shook it swaggeringly.